Preference Model
Member of Technical Staff - ML Platform Engineer
$180K–$350K
San Francisco · On-site · Staff · Full-time · Posted Oct 27, 2026
- Experience
- 5+ years; no alternatives stated.
- Visa sponsorship
- Visa sponsorship and relocation support available.
About the job
Own the platform and integrations that package, validate, scale, and deliver reinforcement-learning environments into partner labs’ training stacks.
What you'll do
- Build partner-lab integrations, container images, and automated delivery checks.
- Build and maintain Karotte, the open-source environment framework, including APIs, releases, and migrations.
- Operate and scale environment build, execution, and grading infrastructure across CPU, GPU, and TPU, with sandboxing and internal developer tools.
What they ask for
- 5+ years building and owning ML platform engineering; AI-related experience is described as ideal.
- Strong software engineering fundamentals and a track record operating production infrastructure or developer platforms.
- Hands-on experience with containers and orchestration (Docker, Kubernetes, or similar) and a major cloud platform (GCP, AWS, or similar).
- Proficiency in Python; working understanding of safely running untrusted code; extensive hands-on use of coding agents (Claude Code, Cursor, Codex, or similar).
Nice to have
- Experience building RL environments, evaluation harnesses, or agent benchmarks.
- Familiarity with RL post-training and inference stacks such as verl, Ray, SkyPilot, vLLM, or SGLang.
Pay
$180K – $350K USD base salary
Tools
Karotte, Docker, Kubernetes, GCP, AWS, Python, Claude Code, Cursor
AI focus
Build and operate ML platform infrastructure for reinforcement-learning environments used in training capable, better-aligned AI models; use coding agents in engineering work.
Model-written summary · the original posting is authoritative